In northern Tanzania, febrile disease caused by the zoonotic pathogens, Leptospira, Coxiella and Brucella spp., account for 11 times more febrile admissions than malaria. These pathogens can infect and be transmitted among a wide range of host species. However, almost nothing is known about transmission patterns among animal hosts, which host species are responsible for transmission to humans, or the key social, economic and behavioral determinants of human disease risk in different agro-ecological settings. This project will integrate several disciplinary approaches, including social behavioral studies, human febrile illness surveillance, and linked human-animal epidemiological studies, to generate data for incorporation into models of human disease risk. These models, together with an understanding of community risk perception and knowledge, will allow us to identify appropriate strategies for disease control and prevention. Intellectual merit: This joint UK-US proposal will be the first integrated study of the impact and social ecology of bacterial zoonoses in Africa. This project brings together medical, veterinary, ecology, and social science research groups that have independently become well-established in Tanzania, but are now working together for the first time. These partnerships provide a unique blend of expertise necessary to conduct the inter-disciplinary research needed to address gaps in knowledge addressing the wide-ranging questions related to zoonoses infection dynamics and disease control. The study will incorporate an interactive methodology with social science approaches both feeding into epidemiological studies as well as building on the outputs from analyses of potential intervention strategies. Broader implications: The scope of the study allows us to investigate disease ecology over a wide range of agro-ecological settings broadly typical of much of rural and peri-urban Africa. Across Africa, increasing attention needs to be given to zoonotic causes of non-differentiated human febrile illness.